Skip to content

Momentum Crashes and Variations to Market Liquidity: summary and key takeaways

The paper tests whether unexpected changes in market liquidity help explain US momentum crashes. It finds that sudden liquidity increases after market declines are tied to large momentum losses, and that this effect absorbs the explanatory power of several known crash predictors.

Butt, H.A.; Virk, N.S. · 2022 · International Journal of Finance & EconomicsSummary updated 20 November 2026
Published2022Journal article
Asset focusEquitiesMomentum
TypeResearchTrend & Momentum
Reading time4 minthis summary

Momentum Crashes and Variations to Market Liquidity at a glance

The question
Do contemporaneous changes in market liquidity explain momentum crashes beyond market rebounds, market variance and the other known predictors?
Dataset & period
US common stocks (NYSE, AMEX, NASDAQ) from CRSP; winner-minus-loser decile momentum returns from Ken French's data library, monthly. Momentum data run from July 1963 to December 2012; regressions use 1965:06-2012:12. Down market states: 90 months, of which 47 saw liquidity rise.
Method
Market illiquidity is the average Amihud price-impact ratio across stocks. Shocks come from an AR(2) filter and are sign-flipped so that a positive shock means liquidity improved. The authors regress momentum, loser and winner returns on liquidity shocks interacted with a down-market dummy, splitting rises from falls. They then add these terms to the Daniel and Moskowitz optionality regression and to regressions with lagged predictors such as market variance, momentum risk and the momentum gap.

Key findings of Momentum Crashes and Variations to Market Liquidity

  • In down market states, momentum returns load on liquidity rises with a coefficient of -9.665 (t-stat -8.92), against 1.233 (t-stat 2.19) for liquidity falls. The effect is strongly asymmetric.
  • The loser portfolio drives this: its coefficient on liquidity rises in down markets is 10.360 (t-stat 8.23). The winner portfolio shows no reliable response to rises.
  • In the 47 down-market months with rising liquidity, losers averaged 3.507% a month and winners 0.702%. Momentum averaged -2.805%.
  • The largest single loser-portfolio return in the extreme-month table was 46.65% in April 2009, while the largest winner return was 4.79%.
  • Once the liquidity-rise term is included, the optionality term of Daniel and Moskowitz loses statistical significance in some specifications. The two variables have a correlation of 40.75%.
  • Lagged market variance, lagged illiquidity and momentum volatility lose significance once the liquidity-rise term is added. Only Huang's momentum gap keeps its predictive power.

Robustness, as reported in the paper

  • A 52-week-high industry momentum strategy, scaled for momentum risk, also shows significant liquidity-rise coefficients, though smaller ones.
  • Results hold after controlling for static and dynamic Fama-French three-factor exposures.
  • Fong-Holden-Trzcinka and zero-return liquidity measures give similar but weaker results. The Pastor-Stambaugh Gamma measure does not give robust results.
  • A common factor from four liquidity measures (second principal component) gives mostly significant results at the 0.01 level; the first component is weaker for some tests.

Caveats and limits

  • Liquidity shocks are measured in the same month as the crash. The analysis explains crashes after the fact and is not a forecast signal (inferred).
  • The sample ends in December 2012, and the full table of extreme months relies on a small number of events (inferred).
  • The paper says not all dimensions of liquidity relate to momentum crashes; the result depends on the liquidity proxy used.
  • No transaction costs or trading-strategy implementation are examined (inferred).

Items marked (inferred) are our reading of the design, not statements by the authors.

Why it matters

The paper offers a liquidity-based account of why momentum drawdowns cluster after market declines. For risk managers, it points to the loser leg's sensitivity to sudden liquidity improvement as a source of tail risk. It also suggests that several known crash predictors may mostly proxy for this one condition.

Questions about this paper

Why do momentum strategies crash after market downturns according to this paper?

The loser leg is more illiquid and reacts far more to sudden liquidity increases than to decreases after declines. Its sharp rebound hurts the short side of momentum.

How large is the asymmetry in the momentum regression?

In down markets the coefficient on liquidity rises is -9.665, versus 1.233 for liquidity falls.

Does the result depend on how liquidity is measured?

Yes. The Amihud measure gives the strongest results, while the Pastor-Stambaugh Gamma measure does not give robust ones.

Where can I read the original paper?

The original is at its publisher: https://doi.org/10.1002/ijfe.2249. We link to the paper and do not host the PDF.

Follow new quant research

Download MacroFinance market data as CSV or Excel, or pull it by API, and test the ideas in these papers yourself. New summaries are added as we review them.

Create a free account
Summaries are written by MacroFinance from the published papers, in our own words, for information only. They are not investment advice and may omit detail. Papers are not hosted here: each page links to the original at its publisher.